The present invention relates to data relaying units which are suitable for use with serial buses and also relates to methods of transmitting data between host and client devices.
In the following description, the term "bus" refers to a data transmission path to which a plurality of devices are connected and whose input/output schedule is managed by a host computer.
With development in information-oriented society today, personal computers have been rapidly coming into wide use in various fields, and in some business organizations, almost every employee has a personal computer on his or her desk. Also, for the purpose of conducting particular operations using a plurality of personal computers, a LAN (Local Area Network) system may be built to share same information and pieces of peripheral equipment such as a printer and a facsimile.
FIG. 11 shows a typical example of a LAN system arranged with star-connecetd token buses, where reference numeral 1 represents a host device or host computer in the form of a personal computer and reference character H represents a hub device connected to the host device to function as a concentrator device for the data buses. Further, reference characters A, B and C represent client devices connected with the hub device H, each of which comprises a personal computer or peripheral equipment. Among these devices, data divided into a plurality of predetermined time units called "time slots" are transmitted via serial buses composed of twisted pair lines. Via the hub device H, the host device 1 controls data transmission between the client devices A-C and between the host device 1 and each of the client devices A-C. That is, the host device 1 control distribution of data flowing through the data buses, and such control is generally known as "bus scheduling".
However, the rates of the data transmissions between the devices are subject to a certain limit depending on the respective performance of interfaces used. Thus, if a large quantity of data are to be transmitted between specific client devices in a time slot, there would arise a problem that the data transmission between other client devices than the above-mentioned specific client devices is limited to a considerable degree. Assuming that a large quantity of data are to be transmitted from the client device A to the client device B in a specific time slot, the host device 1 performs the bus scheduling as shown in FIG. 12. Most part of this specific time slot is occupied by the data transmission period T1 from the client device A to the client device B and the data transmission between the host device 1 and the other client device C is significantly limited because it can be conducted only at the remaining time T2 of the time slot.